47 The effects of human activity and environmental variability on long-term changes in Adelie penguin populations at Palmer Station, Antarctica

DONNA L. PATTERSONI, ANDREA 1. EASTER-PILCHER2 & WILLIAM R. FRASERI IPolar Oceans Research Group, PO Box 368, Sheridan, Montana, USA,' 2University a/Montana - Western. 710 South Atlantic, WlvfC 93, Dillon, Montana, USA e-mail: patterdo@3rivers.net

ABSTRACT

To assess whether human activities due to tourism were negatively impacting Adelie penguins (Pygoscelis adeliae), we compared long-term population trends at visited and control sites on Torgersen Island considering underlying factors associated with environmental variability. To this end, a hillshade model of Torgersen Island was developed; linear regression and discriminant function analyses were used to examine breeding population/landscape interactions.

Results suggest that variability in population trends on Torgersen Island are forced primarily by colony aspect and colony area. Colonies with south-facing aspects are decreasing faster than colonies with north-facing aspects. Smaller colonies are decreasing faster than larger colonies. Both trends are likely due to interactions between the e1fects of increased snow deposition and decreasing egg and/or chick survival due to predation and t1ooding.

To examine human int1uences, subsequent analyses were standardised by pairing Adelie penguin colonies according to area and aspect on the visited and control sides of Torgersen Island. Tourism appears to have no detectable impact on Adelie penguin breeding popUlation size or breeding success; comparisons between population trends in visited and control sides of the island were either not significant or inconsistent with site-specific tourist visitation patterns.

Key Words: Antarctica, Palmer Station, Adelie penguin, human impacts, tourism, landscape effects, population trends

301

INTRODUCTION

Human activity, and tourism in particular, has steadily increased on the Antarctic Peninsula over the past 25 years (Enzenbacher 1992, Cobley et al. 2000). The potential effects that tourism may have on wildlife populations have thus become an impOliant conservation issue that in Antarctica is uniquely magnified because the activities of both humans and wildlife tend to converge on a small fraction of the coastal landscape that is free of ice. This obligatory link between penguins, proximity to marine resources and terrestrial nesting habitat may further amplifY interactions with humans. As a result, the spatial scales over which wildlife-tourist interactions occur tend to be relatively small, but may involve repeated visits to the same sites by hundreds to thousands oftourists over the course of a single season. One concern, therefore, is that the intense human activity associated with these sites will negatively impact wildlife populations.

Coincident with the increasing concern about possible human impacts on wildlife populations is a growing controversy regarding cause and effect. Although some studies have concluded that human activity has adverse effects on wildlife (Reid 1968, Culik et at. 1990, Wilson et at. 1990, Woehler et al. 1994, Giese 1996), other studies do not support these conclusions. A study by Fraser & Patterson (1997), for example, demonstrated that there was no correlation between long­term Adelie penguin popUlation changes in the vicinity of Palmer Station, western Antarctic Peninsula, and the human use histories of area rookeries. Similar findings are described in a study by Cobley et aI. (2000) on tourist impacts at Port Lockroy, western Antarctic Peninsula. This study demonstrated that Gentoo penguins (Pygoscelis papua) not only increased dramatically in the area despite the presence of thousands of annual visitors, but also exhibited no signiticant differences in reproductive success between visited and unvisited control sites. Comparable results have also been reported by Stonehouse (1965) and Taylor et al. (1990), who observed that long­term variability in Adelie penguin breeding populations followed

Antarctic Biology in a Global Context, pp. 301-307. Edited bv A.H.L. Huiskes, Jtv.Jtv.c. Gieskes,1. Ro=ema, R.ML. Schorno, S.M van der T1es & W1. Wolff. ©2003 Backhuys Publishers, Leiden, The Netherlands.

302

/ / i !

i

, ~ ,,}~

n

~ Q

Antarctic

""'~""

! """I

l i

/ Fig. 1. Antarctic Peninsula and Anvers Island

similar trends in areas where human activity was prevalent and where it was absent.

To the extent that cause and effect remains a controversial issue in human impact studies, there is an emerging consensus that the issue

D.L. PATTERSON ET AL.

will not be adequately resolved until we further our understanding of the underlying factors responsible for natural population variability (Woehler et ai. 1994). Indeed, recent studies that adopted this approach have shown that a significant, but previously unrecognized, source of population variability in Adelie penguins is due to interactions between breeding habitat geomorphology and changing patterns of snow deposition (Fraser & Patterson 1997, Patterson 2001). These findings are directly relevant to human impact studies because the implication is that by isolating this natural "landscape" effect on population variability, it may be possible to tease out effects that might be induced by tourism. This landscape effect is linked conceptually to Pulliam's (1988) work on habitat-specific demography in that it utilizes the idea that breeding habitat quality can drive local-scale changes in populations that are independent of factors such as human activity.

The main objectives of this study were to develop and apply a landscape model to examine whether the effects of tourist activities on population trends could be detected above the natural variability induced by habitat geomorphology (cf. Fraser & Patterson 1997). We systematically examined and identified habitat specific features linked to breeding success. The effects of human visitation patterns on Adelie penguin populations were then considered within the context of this landscape perspective. Palmer Station (Fig. 1) has been a popular tourist destination for nearly 30 years, and historically has been divided into areas where visits by tourists are permitted and areas where they are not (Fig. 2). This provided an ideal experimental setting to examine Ad6lie penguin population responses with and without the addition of tourist activities.

Fig. 2. Aerial photograph of Torgersen Island illustrating the division between visited and unvisited Adelie penguin colonies

THE EFFECTS OF HUMAN ACTIVITY AND ENVIRONMENTAL VARIABILITY ON CHANGES IN ADEllE PENGUINS AT PALMER STATION 303

'til C!J

'S ~ u € 0

6-

~ be

~ ~

I:X!l

METHODS

Study Area

The study was conducted on Torgersen Island (64°46'S, 64°04'W, Fig. 2), where tourists have been visiting a large Adelie penguin rookery (ca. 8000 breeding pairs) for nearly three decades. Torgersen Island is 500m in diameter, relatively flat (maximum height above mean sea level is 17m), and is bisected by a low-lying, east-west ridge that roughly divides the island into areas of predominantly north or south aspects.

To examine Torgersen Island's terrestrial habitat, a Digital Terrain Model CDTM) was developed for the island using low-altitude (450m) aerial photographs taken during the 1998/99 season. The aerial photographs were geo-rectified using static ground GPS surveys (Sanchez 1999); the resulting DTM was then used to create a hillshade model of Torgersen Island. The hillshade model shown in Figure 3 illustrates a gradient of habitat quality with a white to black scale indicating areas on the island receiving the least and most amounts of snow, respectively. We used this model to examine interactions between

~

Fig. 3. Hillshade model of Torgersen Island. The coloured areas represent Adelie penguin colonies. The transition from light to black shading represents gradients in landscape aspect; lighter areas are north-facing and are typically scoured free of snow by predominant winds

8000~-------------------------------------------------------------------------r1400

II ;~ ~ ~

7o00+-----~~~------~----------------------------------------__________ ~

6000 Y = -237.73x + 7147.1

r2 = 0.7699

5000

~ ~ . ~ ~ ~.

~

~ ~

4000 ~ ....

~ ~ y = -79.661x + 1364.3 ....... ~

~

~ ... r2= 0.9115 lOot ~ ...

3000

2000~-------------------------------------------------------------------1

1000

--'-North Side

• •• SouthSide

---Linear (North Side)

- ••••. Linear (South Side)

1200

1000

800

600

400

200

0L-----~----~----~----_r----~----~-----~----~----~----~0 89 90 91 92 93 94 95 96 97 98

Year

'til (lJ

'S .sa 0 U .£i

~ .~ ~ be

] ...

I:X!l

Fig. 4. Adelie Penguin Breeding Population Trends on the North and South Sides of Torge~sen Island. Not~ that two y-axes ar; used to rep:esent breeding populations on the north and south sides of Torgersen Island. Over the ensuing lO-year penod, north coloUles decreased by 32Yct versus 58Yct for the south colonies. The slopes of the lines are significantly different (p=O.00383, F1,16=l1.4l5). I

304

breeding habitat geomorphology, patterns of snow deposition and penguin population trends.

The breeding population trends analysed includes 10 years (1989-1998), and is based on colony censuses performed in accordance with standardised methods (CCAMLR 1992) during the peak egg-laying period. Five years of breeding success data (1993-1998) supplement the long-term colony censuses. At the beginning ofthe study in 1989, there were 23 Adelie penguin colonies ranging in size from 30 to 1400 breeding pairs. Tourist access was limited to 12 of these colonies (3310 breeding pairs), leaving 11 colonies (4625 breeding pairs) as unvisited control sites (Fig. 2).

Statistical Analyses Statistical analyses were done using Statistica 99 (Statistica 1999) and Prism 3.02 (GraphPad 2000). Given the dichotomous nature of the data (tourist-visited vs. control-unvisited, north aspect vs. south aspect), linear discriminant function analyses were used to evaluate the predictability of group membership. Specifically, discriminant analysis was used to determine whether colony-specific habitat components (colony aspect, colony area and colony shape) could be used to discriminate between colonies undergoing high population decline (>25%) and those colonies undergoing low population decline «25%) during 1989-1998. Discriminant analysis was also used to evaluate the ability of colony-specific breeding biology parameters (change in number of breeding pairs, percent decrease in breeding pairs per colony, percent decrease in the number of chicks produced per colony, and the number of chicks creched/pair/colony) to discriminate between the tourist-visited and control sides of Torgersen Island. Linear regression analyses were used to assess whether breeding biology parameters supported the model results generated by the discriminant function analyses. Specifically, linear regression analyses were used to examine whether colony-specific habitat components were significant predictors of breeding population parameters (e.g., breeding population size, breeding success, nest loss and chick loss) based on the 1993-1998 breeding data. Linear regression analysis was also used to examine the relationship between snow pack persisting into late November (the peak egg-laying period for Adelie penguins) and the number of chicks creched per colony.

RESULTS

The results of this study suggest that colony aspect and colony area are significant predictors of change in Adelie penguin populations on Torgersen Island. Rates of seasonal snow deposition and accumulation are more pronounced on Torgersen Island's south side as the hillshade model (Fig. 3) clearly demonstrates. Colonies on the south side of Torgersen Island's east-west bisecting ridge (see Fig. 3) decreased at

D.L. PATTERSON ET AL.

Table 1. The numbers and percent of colonies that were correctly classified by the linear DFA into high colony population decrease (>25%) and low colony population decrease «25%) categories (between 1989/90 and 1998/99). The three discriminating variables were colony aspect, colony area and colony shape (perimeter/area ratio). Colony shape did not contribute significantly to the discriminating power of this model (partial lambda = 0.812).

Group Percent correct

Low «25%) 85.71 High (>25%) 100.00 Total 95.65

Colonies classified into

Low High Total

6

0 6

16 17

7

16 23

a rate nearly two times that of colonies situated on the north side (58% vs. 32%, Fig. 4). The difference in the slopes of the regression lines was highly significant (p=0.00383, Fl,l6=11.415).

Fraser & Patterson (1997) speculated that egg and/or chick mortality may need to be considered within the context of colony-specific features, such as size or aspect, to better understand the underlying mechanisms driving natural variability in these populations. A 3-variable discriminant analysis of colony aspect, colony area and colony shape was highly effective in classifYing high and low percentage decline colonies (Table I). Colony aspect and colony area were the strongest predictor variables, while colony shape did not contribute substantially to the discriminatory power of the model. Only one of the 23 colonies was misclassified by the model, thus reinforcing the importance of habitat-specific features in predicting colony-scale population trends. Linear regression analyses strongly supported the discriminant model results presented in Table 1. Colony aspect, colony area and snow depth in November were identified as significant predictors of several breeding success parameters (Table 2).

To look for possible effects due to human activity, subsets of colonies were matched by size and aspect on the visited and control sides of Torgersen Island. Population trends are compared in Figs. 5 & 6. The slopes of linear regressions for north-facing colonies paired by area were not significantly different between visited and unvisited colonies (Fig. 5). However, the slopes of the linear regressions for south-facing colonies were significantly different between visited and unvisited colonies (Fig. 6). Although this may imply a tourism effect, it is worth noting that all 4 colonies representing the visited side of the island were in the areas least attended by tourists. This observation suggests that other influences may be affecting these relatively isolated, yet tourist-accessible, south-facing colonies.

Table 2. Coefficients of determination from linear regression analyses assessing the predictive ability of colony-specific habitat parameters on various reproductive parameters of Adelie penguins.

Co/any aspect Long-Term Breeding Population Change Breeding Population Decrease (percent decrease) excluding a single outlier

identified by residual analysis. Chick Loss (Percent Change) Colony Size (Breeding Pairs)

Colony area Colony-specific Egg Loss at RSGs, excluding a single outlier

identified by residual analysis.

Snow depth Breeding Success (total chicks/total breeding pairs) by Maximum Snow

depth in Late November excluding a single outlier identitied by residual analysis.

r2 p-value Analysis Years

0.36 p<0.003 1989/90 - 1998/99 0.50 p<0.0007 1993/94 - 1995/96

0.38 p<0.004 1994/95 - 1998/99 0.47 p<0.0006 1994/95 - 1998/99

0.75 p<0.0002 1993/94 - 1995/96

0.78 p<O.OOl 1995/96

THE EFFECTS OF HUMAN ACTIVITY AND ENVIRONMENTAL VARIABILITY ON CHANGES IN ADELlE PENGUINS AT PALMER STATION 305

3000r---------------------__________________________________________ ~

2500

• ..

2000

Y'" -60.521x + 2171.3

.~ r2=0.8499

~ bO

1500 -'. ~

, \ ", '-. (II .. . . . •• '"

(II

'" .. ~ Y = -99.77x + 2385.1 ...

r2 = 0.6552

1000

500 - - EcontrOl l .... Visited

---Linear (Control)

- -~~ Linear (Visited) I

O-~----~------r_----~----~------~----~------~----_r----~----~ 89 90 91 92 93 94 95 96 97 98

Year

Fig. 5. Adelie Penguin Breeding Population Trends on the Control and Visited Sides of North-Facing Colonies on Torgersen Island. The trends are based on 5 pairs of colonies that were matched by size at the start of the study (1989). The slopes of the lines are not signiticantly different (p=O.202, F 1.17=2. 094 ).

700.---------------------------------------------------------------------~

500 -~--~---------------------~

200

t' Control I ..... Visited

---Linear (Control)

- - - - - -Lin." (viSite~ 100 -

y = -48.327x + 689.6 r2= 0.9258

" " '. ,

y = -31.333x + 674.73 r2=0.8707

i..~' .~ •.. -_-.-.------------

.....

oL-____ ~------~--__ _r----~------~----_T------r-----~----~----~ 89 90 91 92 93 94 95 96 97 98

Year

Fig. 6. Adelie penguin Breeding Population Trends on the Control and Visited Sides of South-Facing Colonies on Torgersen Island. The trends are based on 4 pairs of colonies that were matched by size at the start of the study (1989). The slopes of the lines are significantly different (p=O.O 180, Fl.lh=6.940).

306

Table 3. The numbers and percent of colonies that were correctly classified as tourist-visited or control colonies. The two discriminating variables were percent decrease in the number of breeding pairs and the number of chicks creched/pair/colony.

Colonies classified into

Group Percent Visited Control Total correct

Tourist (visited) 83.33 10 2 12 Control (unvisited) 63.64 4 7 11 Total 73.91 14 9 23

A 4-variable linear discriminant analysis of change in the number of breeding pairs, percent decrease in breeding pairs per colony, percent decrease in the number of chicks produced per colony, number of chicks creched per colony was unable to adequately discriminate between tourist-visited and control (unvisited) colonies. Six of the 23 (26%) colonies were misclassitied. An examination of the relative magnitudes of the standardized discriminant function coefficients indicated that the variables percent decrease in the number of breeding pairs and the number of chicks creched/pair/colony may have relatively greater predictive power for classifying colonies as tourist or control colonies. However, the discriminatory power ofthis 2-variable model did not improve. Table 3 shows the resulting classitication. Two of the 12 (17%) visited-side colonies and 4 of the 11 (36%) control colonies were misc\assitied, suggesting that tourist activities did not impact these demographic parameters enough to discern group membership.

DISCUSSION

The Adelie penguin is the only Pygoscelid species with a circumpolar distribution and a breeding range that spans nearly 20 degrees in latitude (Woehler 1993). Adelie penguins thus range over an area that varies extensively in the quality and availability of breeding habitat, marine resources and exposure to other wildlife and human activity. Until recently, however, differences in Adelie penguin population trends were explained primarily by variability in the marine environment (e.g., Fraser et af. 1992, Smith et al. 1999). Evidence that these explanations could not account for the differences in trends observed in some local popUlations, however, led Fraser & Patterson (1997) to propose that demographic int1uences involved a minimum oftwo scales of processes. Variability in the marine environment was the most likely explanation for regional-scale trends because the processes would encompass metapopulation scales (i.e., through changes in climate, sea ice or food web processes; Whitehead et af. 1990, Blackburn et af. 1991, Fraser et af. 1992, Smith et af. 1999). Variability in nest site characteristics associated with the terrestrial breeding habitat, on the other hand, was the most likely explanation for differences in local-scale trends because the int1uences only encompassed specific populations (i.e., imposed through mechanisms such as the earlier described landscape eftect; Fraser & Patterson 1997, Fraser et at in prep.). Although this marine/terrestrial model has now emerged as a more robust explanation of demographic variability in Adelie penguins (ct: Fraser & Trivelpiece 1996, Smith et al. 1999), aspects of this model remained untested and its potential application for detecting and interpreting the eftects of human activity had not been addressed. Our results are discussed tl'om this perspective.

This study supports the conclusion that landscape effects appear to have a greater influence on Adelie penguin breeding popUlation trends than human presence on Torgersen Island. In this study, exposure to tourist activities cannot account for all of the observed changes in

D.L. PATTERSON ET AL.

Adelie penguin populations. Stronger predictors of population change included features associated with the geomorphology of the landscape on which penguins colonies were located, and in particular possible interactions between colony aspect, colony area and snow depth (Tables 1~3; Figs. 4-6). To conclude that tourists have no impact on Adelie penguins on Torgersen Island, however, may at this point be premature. This is suggested by Fig. 6, which shows that breeding population decreases in colonies with south facing aspects were larger in the visited areas (69.3% vs. 47.5%). A possible explanation is that poor breeding habitat combined with tourist visits engenders colony-specific cumulative etfects. Another explanation is that the activities of researchers may also be involved (Patterson 2001); this is the subject of current, on-going investigations.

Antarctic tourism is expected to grow substantially in the foreseeable future; hence the conclusion that there are no tourism effects on Adelie penguin demography on Torgersen Island must be placed in its proper context. This conclusion does not imply that tourists aren't having an impact, but rather that none could be detected above the natural variability in the system. National Science Foundation (NSF) management guidelines for Torgersen Island (see Patterson 200 1) limit human contact with Adelie penguins during critical stages in their breeding chronology, which may aftord some protection during the early breeding season. Indeed, few other tourist destinations are as carefully regulated. Given the continued popularity of Antarctic tourism, and considering the results of this and other studies, management guidelines should consider the timing of visits, as well as avoiding small or isolated colonies that may be disproportionately vulnerable to environmental or human int1uences (Tenaza 1971, Patterson 2001). The possibility that Adelie penguins habituate to human visitation should not be discounted, as suggested for other penguin species in the presence of human activity (see van Heezik & Seddon 1990, Cobley et al. 2000). For this reason, managers may suggest that visits be conducted in areas that have consistently received some level of tourism, rather than concentrating on 'expedition touring', in which the focus is to visit new colonies and retreat into areas rarely visited. Further research to evaluate the relationship between tourism and Antarctic penguins, or other wildlife popUlations, should take a cautious approach to the selection of a study site and to their understanding of what influences have shaped and continue to drive the populations on local and regional scales. As summarised in Fraser & Patterson (1997), an understanding of the conditions surrounding human presence is vital to discriminating between human and environmental int1uences.

ACKNOWLEDGEMENTS

We thank the many Held assistants who have helped with the data collection at Palmer Station tor years of dedicated and outstanding fieldwork. We also thank the Antarctic Support Associates staff of Palmer Station and the RlV Polar Duke for logistical support and assistance beyond the typical call of duty. Data from the Palmer LTER data archive were supported by Office of Polar Programs, NSF (OPP-96-32763). This research was funded by National Science Foundation grants to Dr. William R. Fraser (OPP-93-20115 and OPP-95-05596). Additional support was provided by National Oceanic and Atmospheric AdministrationlNational Marine Fisheries Service contracts to Dr. William R. Fraser.

REFERENCES

Blackburn, N. Taylor, R.H. & Wilson, P.R. 1991. An interpretation of the growth of the Adelie penguin rookery at Cape Royds, 1955-1990. New Zealand J. Eco!. 15: 117-121

CCAMLR. 1992. CCAMLR ecosystem monitoring program (CEMP).

---~-------------=------------------,"",,,,,,,,-",,----,,,,----------'iiilill!ilflill~

THE EFFECTS OF HUMAN ACTIVITY AND ENVIRONMENTAL VARIABILITY ON CHANGES IN ADEllE PENGUINS AT PALMER STATION 307

Standard methods for monitoring studies. Scientific Committee for the Conservation of Antarctic Marine Living Resources, Hobart, Tasmania

Cobley, N.D., Shears, 1.R. & Downie, R.H. 2000. The impact oftour­ists on gentoo penguins at Port Lockroy, Antarctic Peninsula. In: Davison, W., Howard-Williams, C. & Broady, P. (eds.) Antarctic Ecosystems: Models for Wider Ecological Understanding, pp 319-323. The Caxton Press, Christchurch

Culik, B., Adelung, D. & Woakes, A.1. 1990. The effects of distur­bance on heart rate and behaviour ofAd6lie penguins (Pygoscelis adeliae) during the breeding season. In:. In: Kerry, K.R., & Hempel, G. (eds.) Antarctic Ecosystems. Ecological Change and Conserva­tion, pp 177-182. Springer, Berlin, Heidelberg, New York

Enzenbacher, 0.1. 1992. Antarctic tourism and environmental con­cerns. Marine Poll. Bull. 25: 258-265

Fraser, W.R. & Patterson, D.L. 1997. Human disturbance and long­term changes in AdeJie penguin popUlations: a natural experiment at Palmer Station, Antarctica. In: Battaglia B., Valencia J. & Walton D. W. H. (eds.) Antarctic Communities: Species, Structure and Sur­vival, pp 445-452. Cambridge University Press, Cambridge

Fraser, W.R. & Trivelpiece, W.Z. 1996. Factors controlling the distri­bution of seabirds: winter-summer heterogeneity in the distribu­tion of Adelie penguin populations. In: Ross, R.M., Hoimann, E.E. & Quetin, L.B. (eds.) Foundations For Ecological Research West of the Antarctic Peninsula, pp 257-272. American Geophysical Union Antarctic Research Series, Washington

Fraser, W.R., Trivelpiece, W.Z., Ainley, D.o. & Trivelpiece, S.o. 1992. Increases in Antarctic penguin popUlations: reduced competition with whales or a loss of sea ice due to environmental warming? Polar BioI. 11: 525-531

Giese, M. 1996. Effects of human activity on Adelie penguin Pygo­scelis adeliae breeding success. BioI. Conserv. 75: 157-164

GraphPad Software. 2000. GraphPad Prism 3.02 Parmelee, D.F. 1992. Antarctic birds: ecological and behavioral ap­

proaches. University of Minnesota Press, Minneapolis, Minnesota Patterson, D.L. 2001. The effects of human activity and environmen­

tal variability on long-term changes in Ad6lie penguin populations at Palmer station, Antarctica. Master's Thesis, Montana State Uni­versity, Bozeman, Montana

Pulliam, H.R. 1988. Sources, sinks and population regulation. Amer. Nat. 132: 652-661

Reid, B. 1968. An interpretation of the age structure and breeding status of an Adelie penguin population. Notornis 15: 193-197

Sanchez, R.D. 1999. GPS and GIS-based data collection and image mapping in the Antarctic Peninsula. USGS OFR 99-402 .

Smith, R.C., Ainley, 0.0., Baker, K.S., Domack, E., EmslIe, S.E., Fraser, B., Kennett, 1., Leventer, A., Mosley-Thompson, E., Stammerjohn, S. & Vernet, M. 1999. Marine ecosystem sensitivity to climate change. Biosci. 49: 393-404

Statistica. 1999. Statistica '99. Statsoft, Inc. Tulsa, OK, USA Stonehouse, B. 1965. Counting Antarctic animals. New Scientist 29:

273-276 Taylor, R.E., Wilson, P.R., Thomas, B.W. 1990. Status and trends of

Adelie penguin populations in the Ross Sea region. Polar Rec. 26:

25-29 Tenaza, R. 1971. Behavior and nesting success relative to nest loca-

tion in Ad6lie penguin populations (Pygoscelis adeliae). The Con-

dor 73: 81-92 van Heezik, Y. & Seddon, P.1. 1990. Effect of human disturbance on

groups of jackass penguins. South Afr. 1. Wildl. Res. 70: 89-93 Whitehead, M.D., Johnstone, 0. W. & Burton, E.R. 1990',~nnual fl~c­

tuations in productivity and breeding success of A~ehe pengu111s and Fulmarine petrels in Prydz Bay, East AntarctIca: In: Kerry, K.R., & Hempel, G. (eds.) Antarctic Ecosystems. EcologIcal Change

and Conservation, pp 214-223. Springer, Berlin, Heidelberg, New York

Wilson, K.-J., Taylor, R.H. & Barton, K.1. 1990. The impact of man on Adelie penguins at Cape Hallett, Antarctica. In: Kerry, K.R. & Hempel, G. (eds.) Antarctic Ecosystems. Ecological Change and Conservation, pp 183-190. Springer, Berlin, Heidelberg, New York

Woehler, E.1. 1993. Distribution and abundance of Antarctic and Sub­antarctic penguins. Scientific Committee on Antarctic Research, Cambridge, 76pp

Woehler, E.1., Penney, R.L., Creet, S.M. & Burton, H.R. 1994. Im­pacts of human visitors on breeding success and long-term popu­lation trends in Ad6lie penguins at Casey, Antarctica. Polar BioI. 14: 269-274

I